Details
-
Bug
-
Status: Open
-
Minor
-
Resolution: Unresolved
-
3.6.1
-
None
-
None
-
2 CentOS Linux Servers
Description
We have a master/slave solr setup. We did some fail over tests with solr regarding replication. We corrupted the master index ( we deleted files ) during runtime. During the next replication of the slave it trasfered the broken index from the master server and during that we got an fileNotFound Exception which stopped the replication. So the slave index was not overwritted and still working. But there is still a temp folder on the disk ( like /data/index.xxxxxxxxxxxxxx/ ). This happened after every replication until the disk was full. So a lot of temp folders was created.